This thesis describes a quadruple-resonance 1H/13C/2H/15N magic-angle spinning nuclear magnetic resonance probe for the study of structure and dynamics of biological macro- molecules. The probe utilizes tuning-tube components as discrete circuit elements. A coaxial design allows specialization, with the high-frequency 1H circuit on a low-voltage modified Alderman-Grant coil and the 13C/2H/15N circuit on a solenoid coil for more efficient detection. The mechanical and circuit design of the probe is presented with a description of the benchtop characterization and some initial experimental results.

Advances in technology and analytical instruments are driving factors in scientific progress, particularly in the area of nuclear magnetic resonance. However, these developments have often not been easily accessible, creating a barrier to collaborative and broadly impactful use, especially in structural biology. My research has focused on innovative approaches to NMR instrumentation design to support characterization of complex biomolecular assemblies such as crystallin hydrogels that form the eye lens or membrane-associating antimicrobial peptides identified from the carnivorous plant Drosera capensis. My work has explored applications of additive manufacturing, which has emerged as a technology that rivals and occasionally surpasses conventional fabrication methods, to the NMR instrumentation design process in terms of design achievability, relative ease of use, availability, and material library. Interest in modifying both purpose-built and commercial probes led to the development of a creative, efficient, and accessible method for NMR transceiver coil fabrication using removable 3D-printed templates. Experimental magnetic-field profiles correlated with the results of theory-driven software simulations, demonstrating the ability to fabricate verifiably unique designs for a variety of experimental applications. This method not only enables coils to be made to specification, but also supports complex designs such as saddle-coils or others that are unable to be achieved by hand such as continually-variable pitch solenoids. Expanding upon this work led to creation of a generalized and completely open-source approach in support of modularity to quickly achieve optimized solenoid transceiver designs based on system-specific or user-defined constraints. Parameter spaces defining suitable variable-pitch solenoids were plotted in an adaptable Python workspace, yielding options that predict improvements over previously published designs. The magnetic field profiles of every viable design were evaluated based on two performance-driven figures of merit in order to identify optimized designs for experimental testing and validation using a recently developed open-source and automated benchtop approach. In order to adequately maintain transceiver coil integrity in mechanically-dynamic systems 3D-printing was further leveraged in collaboration with 3M to produce the first of their kind polytetrafluoroethylene parts at dimensions previously unable to be fabricated. Implementation of this work and complimentary open-access approaches has tremendous promise to transform the field of structural biology by expanding participation of both novices and experts alike in use or development of modular components optimized for specific challenges.

Crystallins are water-soluble proteins that are necessary for focusing light on the retina. In mammalian lenses, there are two classes of crystallins; α-crystallins (molecular chaperone proteins), and βγ-crystallins (structural proteins). My research has focused on $\gamma$S-crystallin, which is the main structural component of the human eye lens cortex and contributes to the lens fiber organization. Present up to 450 mg/mL in the human lens, γS-crystallin depends on its long-term ability to remain stable and retain a high refractive index. The G18V variant of γS-crystallin (γS-G18V), associated with hereditary childhood-onset cataract, shows decreased stability and increased aggregate formation. My work explores the intermolecular interactions contributing to the increased aggregation propensity of γS-G18V relative to wild-type (γS-WT). By titration of a hydrophobic chemical probe (ANS) and by titration of tripeptides (a library composed of the crystallin sequence), residue-specific binding was observed via NMR chemical shift perturbations (CSP). Protein-protein interactions have also been measured for the direct determination of the second virial coefficient, which describes the intermolecular interactions as being either repulsive or attractive. The second virial coefficient for lysozyme under physiological conditions will give better insight into how the individual molecules are interacting with one another in solution. The control experiments with lysozyme will be applied to both γS-WT and a deamidated variant, N15D, to investigate how prone the crystallins are to aggregation. Additionally, in order to investigate the functionality of crystallins from aquatic species, the biophysical characterization of the J2-crystallin from box jellyfish (Tripedalia cystophora) will be presented. J2-crystallin is an excellent example of convergent evolution developing a protein to perform the refractive function needed in the eye lenses and thus a deeper understanding of how these crystallins are related to one another. Lastly, the refractive index increment (dn/dC) of various crystallins from aquatic and terrestrial species will be reported to fully understand the functionality of the eye lens.

This thesis describes a quadruple-resonance 1H/13C/2H/15N magic-angle spinning nuclear magnetic resonance probe for the study of structure and dynamics of biological macro- molecules. The probe utilizes tuning-tube components for discrete circuit elements. A coax- ial design allows specialization, with the high-frequency 1H circuit on a low-voltage modified Alderman-Grant coil and the 13C/2H/15N circuit on a solenoid coil for more efficient detec- tion. The mechanical and circuit design of the probe is presented with a description of the benchtop characterization and some initial experimental results.

Clathrate hydrates are a class of compounds in which small guest molecules, such as methane, ethane, tetrahydrofuran, etc. are encapsulated within a hydrogen-bonded host water cages. In the oil industry, vast deposits of methane hydrates in deep ocean and permafrost have received great interest as a potential energy source. However, methane hydrate formation can potentially block the transmission pipelines in deep see drilling which may lead to catastrophic economic losses from the flow assurance failures. Therefore, understanding the fundamental mechanisms of gas hydrates formation, decomposition and inhibition is the key for their safe and economical use as a potential energy source and natural gas storage. The effect of low concentrations of methanol on the clathrate hydrate formation kinetics were studied in this thesis. The catalytic effects from methanol were observed on propane hydrate formation but inhibitory effects were found in fluoromethane hydrate formation. We postulate that the catalytic activity of methanol is dependent on the nature of the guest gas itself. Moreover, one of the fascinating properties of the clathrate hydrate structure is that the water molecules of host lattice completely satisfy the hydrogen-bonding requirement, so the interactions between the guests and host cages are mainly due to van der Waals forces rather than hydrogen bondings. Within the cages, the trapped guest molecules have limited translational mobility but may retain rotational and vibrational freedom. From the standpoint of physical chemistry, the dynamics of the guest and host molecules is an excellent material to study and the motion of the guest molecules within the cages is not well understood. In this thesis, the dynamics of tetrahydrofuran (THF) and cyclopentane (CP) guests in the hydrate cages above 200 K were investigated by magic angle spinning (MAS) solid-state NMR. The barrier to guest motion of CP is much lower than THF indicating the existence of hydrogen bonding interactions between THF guest and cage water molecules above 200 K.

This thesis examines the biophysical properties of beta/gamma-crystallin proteins with a specific focus on how exogenous factors such as UV and divalent cations drive the aggregation of eye lens gammaS-crystallin. Beta/gamma-crystallin superfamily proteins are found in archaeal, bacterial, and eukaryotic species, with varying functionalities that represent evolutionary adaptation. In this work the eye lens gammaS-crystallin is shown to have orthogonal behavior to Ciona intestinalis beta/gamma-crystallin (Ci-b/g) – a calcium binding protein – in the presence of various divalent cations. Ci-b/g coordinates divalent cations at the same site, increasing its thermal stability, whereas cystine coordination to some transition metal divalent cations induces aggregation in gammaS-crystallin. Unlike other transitional metals, copper induced aggregation of gammaS is buffered by the presence of free, solvent accessible cysteines. Copper-driven aggregation differs from other underlying causes such as UV or missense point mutations and may be attributed to a combination of different factors. Notably, copper is able to catalyze intermolecular disulfide bond formation as well as facilitate radical crosslinking. Reductions in beta-sheet content also occur in aggregated species, further suggesting that hydrophobic interactions may drive aggregation. Another physiologically relevant underlying mechanism, UV radiation, was used to induced aggregates for comparative measurements with acid-induced fibrils that exhibit amyloid character. The two groups are distinct in their biophysical characterization. High levels of turbidity but low thioflavin T binding for UV-induced aggregates while acid-induced fibrils exhibit the opposite characteristics. In both cases, it is notable that mutations compromising stability exacerbate the observable characterization. These results suggest that gamma-crystallin aggregates resulting from accumulated UV exposure are unlikely to contain fibrillar character.

βγ-crystallins are structural proteins in the eye lens that refract light to produce an image. These proteins are found at extremely high concentrations in the lens and must remain sol- uble and transparent. If their solubility is perturbed due to a mutation, chemical damage, or a disturbance in metal ion homeostasis, then cataract can develop and cloud the eye lens. In order to understand more about the evolutionary origin of the structure and aggrega- tion of these proteins in the human eye lens, the structure and metal binding properties of an ancestral Ca2+ binding βγ-crystallin found in the Ciona intestinalis tunicate were investigated using biophysical characterization techniques such as fluorescence, CD, ITC, DLS and solution-state NMR. It was found that the tunicate crystallin has a conformational change upon addition of Ca2+ and other divalent cations. This protein is highly stabilized both chemically and thermally upon binding to Ca2+ and it binds more strongly than has been previously reported for other Ca2+ binding βγ-crystallins. It was also found that this crystallin interacts with other divalent cations, all of which thermally stabilize the protein. Although divalent metal ions raise the melting temperature of the tunicate crystallin, aggre- gation is induced by the interaction with some of these metal ions well below the melting temperature. Several solution-state NMR techniques were developed in order to solve the structure of this tunicate βγ-crystallin and other related biomolecules in the presence of divalent metal ions. One of the methods uses DLPC/DHPC bicelles, a low-temperature alignment medium, to obtain long range distance and angular restraints of dynamic and heat sensitive biomolecules in the presence of divalent metal ions. These restraints are de- termined via measurement of residual dipolar couplings, where J-couplings of a biomolecule in isotropic solution are compared with the J+D values for the biomolecule in the presence of an alignment medium. If there is a change in this peak splitting, then alignment of the biomolecule is occuring. Another technique involves using a small molecular probe dCDP and 31P NMR to determine the macromolecule-free and macromolecule-bound divalent metal ion concentrations in NMR samples out of a total amount of divalent metal ions put into the system.

This thesis explores the eye lens proteins from the Antarctic toothfish Dissostichus mawsoni and how they have evolved to the subfreezing temperatures of their environment. This includes showing that toothfish γS1- and γS2-crystallins are less stable than their homologous human counterparts and working toward solving their structures using solution-state NMR. By making structural and biophysical comparisons, inferences can be made about how these crystallins have become cold adapted. Another unique adaptation of D. mawsoni is the ability to completely resist cold cataract, a liquid-liquid phase separation of the proteins. By studying γM-crystallins from D. mawsoni that are susceptible to phase separation and performing site specific mutagenesis it was discovered the temperature of phase separation could be controlled by simply swapping between lysine and arginine residues. These results hint at hydration effects and salt bridges as being a major factor influencing the crystallin’s propensity to self associate into a separate liquid phase. To further characterize these proteins their functional role of providing refractive power was measured. The results of these measurements demonstrated that all the crystallins tested had a refractive index much higher than would be predicted by their amino acid composition. This suggests that protein conformation has a large impact on protein refractivity and the assumed models used to predict protein refractivity must be approached much more carefully.

Also described is the D1-PSI discovered in the genome of the carnivorous plant Drosera capensis. As a saposin-like protein, it has demonstrated the ability to interact with membranes in a way that can exhibit anti-microbial growth. Results show that this PSI is able to disrupt membranes, but seems to lack bias for which lipid head groups it interacts with in the context of a stable lipoprotein complex. To better understand what is structurally happening to the D1-PSI while interacting with a membrane, solid-state NMR experiments are ongoing. Such information can inform the mechanism by which D1-PSI is capable of disrupting membranes.

As genomic repositories increasingly grow with a variety of data from a multitude of organisms, the need to approach extracting and interpreting data also becomes increasingly difficult. Recent advances in protein annotation and structure prediction have improved, however the variety and sheer amount of data requires unique approaches from multiple different disciplines. Bioinformatics yields important functional sequence information and classification. Molecular dynamics (MD) simulation allows for the interrogation of biochemical systems at the atomistic level. Combined with machine learning, these disciplines can be equipped to investigate the complex functions and relationships of proteins within the current abundant genomic landscape.

The objective of this dissertation is to outline complementary methodologies from various fields - bioinformatics, molecular dynamics simulation, and machine learning - that together, can investigate vast genomic repositories, functional protein data.

Aim 1: The development of the bioinformatics and in silico maturation pipeline consists of gene annotation, MD simulation to equilibrate predicted proteins, and statistical methods adopted from graph theory in collaboration with the Butts lab. Proteins can be represented in graph theoretic terms allowing for the exploration of diverse protein structural features.

Aim 2: Molecular dynamics simulation gives rise to atomic level details of complex systems. A variety of protein systems - HIV Rev, short intrinsically disordered peptides, STXPB4, YAP-1 WW domain - explored are intrinsically disordered. MD simulations were used to simulate the complexities and difficulties encountered within these proteins as well as plant metabolic proteins.

Aim 3: After the aforementioned bioinformatics pipeline and \textit{in silico} molecular dynamics-based maturation of predicted proteins, methods to extract useful atomistic information from coarse protein structure networks (PSNs) were developed. A multi-layer perceptron was used to essentially upscale coarse PSNs into atomistic models. The significance of this technique permits for the simulation of coarse PSNs, and the exploration of complex protein structural conformations.

Proteins are fundamental building blocks of life: understanding protein structure, function, aggregation, and degradation is, therefore, one of the central questions in biology. My work investigates protein aggregation and degradation through computational modeling, protein structure network analysis, and experimental verification.

One theme of my work is the discovery of new enzymes from the carnivorous plant, \Drosera capensis (D. capensis). With the ever-expanding genomic data, it is imperative to swiftly move from raw genomic data to chemical results. Using the ``target selection pipeline'' that we invented, in silico protein structures can be predicted rapidly, to direct the subsequent experimental characterization of the promising candidates. Subsequent network analysis predicts interesting protein properties such as potential enzyme activity, enzyme specificity, and the functional pH range, aiding the selection of functionally useful proteins for experimental characterization. This approach illustrates a generally applicable way to leverage the wealth of information provided by whole genome shotgun sequencing for proteomics. Computational techniques, despite their limitations, are now powerful enough to allow potentially useful proteins to be identified directly from the genome and filtered for strong indicators of biochemical function. So far, this work has resulted in three publications including proteases, chitinases, and esterase/lipases.

The protease resistance of amyloid fibrils and their central role in more than 40 human diseases, including Alzheimer's, makes them an attractive target to test the activity of new proteases from \textit{D.capensis}. To advance and streamline scientific discovery related to amyloid fibrils, it was crucial to have a standardized nomenclature. With collaborators, I introduced a systematic approach to the nomenclature of fibril topology using graph theoretic concepts to abstract the structure. The scheme encompasses all amyloid fibrils currently in the Protein Data Bank (PDB), and can be easily extended to accommodate newer discoveries. The work also showed that the vast majority of known fibril structures fall into just three topological categories, something that was previously unnoticed. My work has improved the discussion of fibril structures by condensing the descriptions of complicated structural features using a set of universal structural motifs.

The other theme of my work includes solving the protein structure of J2 crystallin, an aggregation resistant protein. J2-crystallin is a novel eye lens protein, highly expressed in Tripedalia cystophora (box jellyfish) and is an interesting target because of its very high stability and water-solubility. Unlike most non-cephalopod invertebrates, box jellyfish have camera-type eyes; therefore, their crystallins present an interesting system from an evolutionary biology perspective, making them an intriguing model system for vertebrates. Interestingly, Basic Local Alignment Search Tool (BLAST) search of J2 in the Protein Data Bank (PDB) found no proteins above 32 \% similarity. Therefore, the structure determination of J2 is not only important from the evolutionary standpoint but also because of the hypothesis that J2 possesses a novel protein fold, due to a lack of known homology. Here, I present the biophysical characterization, and solution-state NMR assignments of J2 crystallin, a previously uncharacterized eye lens protein, addressing the interplay of sequence, structure and function in the eye lens crystallins.